🔍 Read the full analysis: Google’s First Data Center Satellite: Poetry, AI, And The Limits Of A Test on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Google says a satellite carrying its data-center-class TPU has completed its first AI response in orbit: a short poem generated by its Gemma 4 model. The test shows that a small AI workload can run in space, but power, heat, radiation, connectivity and launch costs remain major barriers to a usable orbital data center.
Google says a satellite carrying one of its data-center-class TPUs generated its first AI response in orbit this week, a short poem produced by the Gemma 4 model as the spacecraft passed over Svalbard. The test is an early milestone for Project Suncatcher, Google’s research effort into whether solar-powered satellites could eventually host AI computing, but it does not show that an orbital data center is practical or economical.
Google reported the response on October 30, describing it as the project’s “Hello World” moment. Engineers sent a prompt asking for a haiku about space during a ground-station pass. Google said the satellite returned: “Blue marble below / Data streams in silent grace / Starlight feeds my core.” The company also said engineers tested space-themed limericks and asked the payload to report its technical status.
The prototype launched on October 1, 2026, on a SpaceX Transporter-18 rideshare mission, alongside more than 100 other spacecraft, according to the source material. Built with satellite company Planet on a standard Planet platform, it carries four TPUs and can generate up to about one kilowatt of power. TechCrunch reported that power and heat constraints mean only one TPU can operate at a time; the chip runs in bursts, with the latest reported tests lasting about 20 minutes.
Google says the satellite spent about three weeks in commissioning, including checks of its host computer and TPU trays and early machine-learning tests. The company described the spacecraft as “operating as expected.” It now plans tests of TPU performance under thermal stress and radiation exposure. Those checks concern the hardware’s ability to operate in orbit; the poem itself is evidence of a successful small workload, not a demonstration of a multi-satellite computing system.
Infographic 01 · What flew and what it did
One satellite, four chips, one haiku
Google’s Project Suncatcher prototype is the first test of whether its AI data center chips can work in orbit. Here is what is on board and how it got there.
- 4 Nov 2025Project Suncatcher announcedResearch post and preprint; plan: two prototypes with Planet by early 2027
- 1 Oct 2026Prototype launchedSpaceX Transporter-18 rideshare, built with Planet; peer-reviewed paper in Joule
- Oct 2026About three weeks of commissioningHost and TPU trays powered up, early machine learning tests
- Week of 26 Oct“Hello World”First prompt to Gemma 4; the haiku came down during a pass over Svalbard, Norway
- NextHeat and radiation tests in orbitHow the TPUs behave under thermal stress and real space radiation
- 2027Twin interconnected satellitesMore power and laser links between satellites (Google, TechCrunch)
Scale check
The satellite vs. one power deal on Earth
890 MW = 890,000 kW, about 890,000 times the prototype (my arithmetic). The prototype bar is drawn at a visible minimum; to scale it would be invisible.
Not the first AI chip in orbit
Starcloud-1, launched November 2025 (CNBC)
- Carried one Nvidia H100, a data center GPU
- Trained a small language model in orbit
- Ran Google’s open model Gemma up there in December 2025
- Google’s claim is narrower: the first response from a data center class TPU in orbit
Sources: Google, 30 Oct 2026 · Google, 1 Oct 2026 · Google Research, 4 Nov 2025 · TechCrunch, 30 Oct 2026 · TechCrunch, 1 Oct 2026 · Google and Constellation, 6 Oct 2026 · CNBC, 10 Dec 2025.
Powered by Thorsten Meyer AI
The Cost of Moving AI Compute
The experiment addresses a real pressure on data-center operators: AI computing requires substantial electricity, and demand for data-center power is drawing increased attention. The source material cites U.S. Department of Energy estimates that data centers used about 4.4% of U.S. electricity in 2023, with the department projecting a range of 6.7% to 12% by 2028. These are national figures, not a measure of Google’s own consumption.
Google’s proposed alternative is to capture solar energy in orbit, where it says panels in a suitable orbit could be up to eight times more productive than on Earth. But stronger sunlight does not by itself make orbital computing cheaper. Google’s research estimates that, at current launch prices, launching the hardware works out to about $14,700 per kilowatt per year. In its lower-launch-cost scenario—$200 per kilogram—the figure falls to about $810. The research says reaching that scenario would require roughly 1,800 Starship launches.
Those are estimates and assumptions, not operating costs measured for a commercial orbital data center. They illustrate why the test matters without resolving the business case: a working chip in orbit is a technical milestone, while a useful, affordable computing service would require far more power, repeated launches and reliable links between spacecraft and Earth.
Products Worth Considering
Apple MacBook Air 15-inch (2026) with M5 chip, 16GB memory, 512GB SSD, Silver
LG UltraGear 32GX870A-B 32-inch 4K OLED Gaming Monitor
LG UltraGear 27G640A-B 27-Inch QHD IPS Gaming Monitor, 300Hz, 1ms, G-SYNC Compatible, FreeSync Premium, DisplayHDR 400
As an affiliate, we earn on qualifying purchases.
From Suncatcher to a Single Satellite
Google introduced Project Suncatcher in a research post and preprint on November 4, 2025. Its concept is a constellation of solar-powered satellites carrying TPUs and connected by laser links. The design aims for nearby spacecraft to share computing work as a cluster, rather than treating each satellite as an isolated computer. Google’s published research sketches clusters of 81 satellites within a radius of about one kilometer at an altitude of roughly 650 kilometers, with neighboring satellites 100 to 200 meters apart.
The project’s timetable and the flight described here are not identical in scale. Google had previously named a two-satellite “learning mission” with Planet as a milestone for early 2027. The October 2026 flight is a smaller prototype; Google says the next milestone, if testing goes well, is two interconnected satellites in 2027. TechCrunch reported those spacecraft are intended to generate more power and test optical links.
The “first” claim also needs a narrow definition. The source material says startup Starcloud launched a satellite with an Nvidia H100 GPU in November 2025, and CNBC later reported that it had trained a small language model and run queries to Google’s Gemma model in orbit. Google’s claim is specifically about an AI response from a data-center-class TPU in orbit, not the first AI model or AI chip operated in space.
““the first-ever AI response from a data center class TPU in orbit””
What the Orbital Test Cannot Show
The response confirms that the prototype could run a small AI task and return its output during a ground-station pass. It does not establish continuous operation, commercial capacity or lower costs than terrestrial data centers. The source material does not give a complete accounting of the satellite’s launch, build, communications and operating costs, nor does it report performance over an extended period.
Key engineering questions remain: how the equipment will shed heat in a vacuum, how it will withstand radiation, how failures could be repaired, and how satellites will maintain high-capacity links with one another and with ground stations. Google’s research proposes close satellite clusters and cites a laboratory laser-link test, but the supplied material does not establish that those links have been demonstrated between the planned spacecraft in orbit. The source also notes about 47,000 tracked objects already in orbit; it does not specify a time or classification breakdown for that count.
The distinction between Google’s test and earlier orbital AI work also remains important when describing the milestone. Starcloud’s reported GPU and model activity predates this flight, while Google’s narrower TPU claim is its own. Broader claims that orbital data centers can ease terrestrial electricity demand or outperform Earth-based facilities remain proposals, not results of this test.
Thermal Tests and Twin Satellites
Google says it will next test the TPU payload under thermal stress and radiation exposure. The outcomes will indicate whether the hardware can keep operating in the conditions it faces in orbit, but the company has not supplied dates or detailed success criteria for each test in the source material.
If the tests proceed as planned, Google’s next stated milestone is a pair of interconnected satellites in 2027. TechCrunch reported that the pair is intended to have more power and to test optical links between spacecraft. That flight would address a larger part of Suncatcher’s concept than the current single-satellite demonstration, though it would still not by itself establish the economics or scale of a commercial orbital data center.
Key Questions
What did Google’s satellite do?
Google says the satellite ran its Gemma 4 model on a TPU and returned a short poem after engineers sent a prompt during a ground-station pass over Svalbard.
Was this the first AI system in space?
Not according to the source material. It says a Starcloud satellite carrying an Nvidia H100 launched in November 2025 and later ran AI workloads. Google’s claim is narrower: its satellite produced a response from a data-center-class TPU in orbit.
Is Google operating a full data center in orbit?
No. The flight described is a small prototype with four TPUs and up to about one kilowatt of power. The reported tests run one TPU at a time in bursts; a multi-satellite computing cluster remains a planned development.
What happens next for Project Suncatcher?
Google says it will test the hardware under thermal stress and radiation exposure. Its next stated milestone, if those tests go well, is a pair of interconnected satellites in 2027 to test optical links.
Source: ThorstenMeyerAI.com
